# Robotic Manipulation on Real-world ALOHA OOD

Source: [Robotic Manipulation on Real-world ALOHA OOD](<https://www.sota2.com/research/sota/robotic-manipulation-on-real-world-aloha-ood>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-05-29T11:41:10.582Z

Evaluation of Robotic Manipulation performance on the Real-world ALOHA Out-of-Distribution (OOD) setting, assessed using metrics related to Color fidelity, Instance recognition, Position accuracy, Background consistency, and Instruction adherence.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Robotic Manipulation |
| Dataset (including any stated split) | Real-world ALOHA OOD |
| Metric | Color Fidelity |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Color Fidelity](<https://www.sota2.com/research/sota/robotic-manipulation-on-real-world-aloha-ood.md?metric=Color%20Fidelity>): Higher is better; unit: PERCENT.

- [Instance Recognition](<https://www.sota2.com/research/sota/robotic-manipulation-on-real-world-aloha-ood.md?metric=Instance%20Recognition>): Higher is better; unit: PERCENT.

- [Position Accuracy](<https://www.sota2.com/research/sota/robotic-manipulation-on-real-world-aloha-ood.md?metric=Position%20Accuracy>): Higher is better; unit: PERCENT.

- [Background Consistency](<https://www.sota2.com/research/sota/robotic-manipulation-on-real-world-aloha-ood.md?metric=Background%20Consistency>): Higher is better; unit: PERCENT.

- [Instruction Adherence](<https://www.sota2.com/research/sota/robotic-manipulation-on-real-world-aloha-ood.md?metric=Instruction%20Adherence>): Higher is better; unit: PERCENT.

- [Average Score](<https://www.sota2.com/research/sota/robotic-manipulation-on-real-world-aloha-ood.md?metric=Average%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Color Fidelity | Result date | Report paper |
| --- | --- | --- | --- |
| Qwen-VLA-aloha | 88.5 | Not recorded | 6a191dbedbd7150506f82a0f |
| π0.5 | 57.7 | Not recorded | 6a191dbedbd7150506f82a0f |
| GR00T N1.6 | 46.2 | Not recorded | 6a191dbedbd7150506f82a0f |
| Qwen-VLA-aloha | 42.3 | Not recorded | 6a191dbedbd7150506f82a0f |

### Qwen-VLA-aloha

Color Fidelity: 88.5. Result ID: 6a197b4cfbc373592c14d352.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a0f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a0f.md>)

Recorded evaluation configuration: {"Pre-training strategy":"fine-tuned from pretrained Qwen-VLA-Base"}. This record does not assert independent reproduction.

### π0.5

Color Fidelity: 57.7. Result ID: 6a197b4cfbc373592c14d350.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a0f.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### GR00T N1.6

Color Fidelity: 46.2. Result ID: 6a197b4cfbc373592c14d34f.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a0f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa779a1e209131cea74489.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Qwen-VLA-aloha

Color Fidelity: 42.3. Result ID: 6a197b4cfbc373592c14d351.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a0f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a0f.md>)

Recorded evaluation configuration: {"Pre-training strategy":"trained from scratch"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.